Spinning Yarns

When I heard that the FCC was considering lifting its ban on cellular telephone usage during commercial airline flights, I was dismayed. I’ve never really bought into the argument that they pose any real risk of disrupting the avionics, but I like the fact that you don’t have to spend the whole time listening to some oily-haired cologne-soaked half drunken businessman in a polyester suit yacking about strategy to some faceless schmuck ten thousand miles away. It’s bad enough when he tries to explain it to you in person; when the other half of the conversation isn’t even in the room, it’s practically intolerable.

Fortunately, the ban remains in place, at least for the moment. But you still encounter lots of people who seem to enjoy having loud conversations on their cell phones in the middle of a public space, and it can be really annoying for anyone else in the vicinity.

Happily, I recently discovered a surprisingly effective technique for getting public cell phone users to take their conversations elsewhere, and I hope that if I tell people about it, it will catch on. Here’s how it works: Get as close as you can to the person who is talking on the phone, without invading their personal space or making them feel threatened. Often, you don’t even have to move, since they’re usually standing right next to you. Now, look right at their face, and listen to their conversation. You shouldn’t frown or growl or be overtly rude; in fact, I find it is best if you smile and nod (and, if necessary, say “mm-hmm” at appropriate points in the dialog). Pretty soon, they will turn away, and if you keep watching and listening, they will usually walk a little ways off and lower their voice a little bit, so as to recover some of their privacy. Success! Do not follow them—this isn’t about assault or intimidation. You’re just breaking the social convention about ignoring strangers when you’re in a public space.

If they look at you, your natural inclination will be to look away, but don’t do that! Instead, look right back at them. Smile, if necessary, to make it clear that your attention is benign, but don’t let them stare you down.

As far as I can tell, the withdrawal response is almost completely automatic, like putting out your hand in response to a proffered handshake. What’s more, this technique doesn’t require interrupting their conversation to ask them to move, yet it ostensibly leaves them with a choice about how to respond—although so far, it seems that everybody chooses to withdraw. You will probably have to adjust the technique for application to cultures outside the United States, particularly where social norms about gaze and attention are more strict, but I suspect you can find some variation of this that will work just about anywhere loud cell phone use is prevalent.

Filed by Michael at 19:46 under Social
4 Comments

How much of your life is contained in a heartbeat? The cycles of that miraculous organ define the essential quanta of human time, sometimes running fast, and sometimes slow, but never silent as long as you live. When you are idle, your heart beats on the order of 60-90 times each minute, on average,¹ so let’s split the difference and say it averages out to about 72 beats per minute over the course of your life. Tick, tick, tock—the heart’s your body’s clock. The U.S. Census estimates the life expectancy of males (at birth) to be 73.8 years, and of females (at birth) to be 79.5 years,² so if we’re willing to play it fast and loose and say that men and women are about half-and-half in the population at birth,³ we can say that the average life expectancy is around 76.65 years. At 72 bpm, that works out to an average of heartbeats in the course of an American human being’s life. I’m pretty damned sure those low-order digits are not significant, so let’s just call that 2.902 billion (thousand million) heartbeats and have done with it.

A computer also has a heartbeat, defined by the pulses of a tiny transistor-based oscillator etched into the surface of the silicon die that forms the substrate for its microprocessor. Just like a human’s, the computer’s heartbeat may vary in speed somewhat as it runs—either by design (e.g, to trade between performance and power consumption) or because of the normal variations of voltage at the input source. But a computer’s heart beats much faster than a human’s, and we typically measure it in cycles-per-second (Hz) rather than beats-per-minute. 72 bpm works out to 1.2 Hz. A typical microprocessor—say, a modern Intel or AMD CPU—will run on the order of two or three billion times faster then a human heart. These days, a typical CPU clock speed is on the order of 2-3 GHz.

Okay, so what? Computers are fast, and people are slow. That’s not news—heck, it’s the reason we built electronic computers in the first place! What’s interesting to me about these figures is just how much faster computers are than we are. To put this into perspective, I propose we define a “subjective minute” as the length of time required to measure out 72 heartbeats. For a human, a subjective minute is basically the same as a “real” minute.⁴ But for a computer, a subjective minute takes much less time—a processor running at 1MHz (the approximate clock speed of the Motorola 65C02 used in the old Atari systems) experiences 72 cycles in 0.000072 seconds. Turn that around, and you might therefore say that the 65C02 experiences approximately 13,889 subjective minutes in the course of each real second. That’s around 231.5 subjective hours or 9.6 subjective days.

Using this mental model, you can get a better sense of how fast a computer really goes, compared to us. Let’s suppose, for the sake of argument, that you’re typing at your computer, and you can touch-type at a rate of 60 words per minute. A standard double-spaced typewritten page of English text is around 250 words, and English prose averages about 5 characters per word, plus one for the spaces in between words. With these assumptions each page you write contains around characters. If we assume you’re typing at a more or less “even” rate, we would expect you to produce a character every 0.173 seconds on average.

Now put yourself in the 1 MHz computer’s shoes: In 0.173 seconds, your heart beats 173,000 times, a total of about 2,403 subjective minutes. That works out to about 40 hours. Now, imagine you were receiving a note from a friend, one character at a time, and you had to wait 40 hours for each one! You can begin to get an idea of how it is that the computer can afford the time to play music, copy files, download porn, and run several other applications while you are busily typing away at your thesis. The simple fact is, you just can’t keep it very busy just by typing. The situation is even worse for the poor 2.8 GHz machine, which is 2,800 times faster than the 1 MHz machine, and therefore experiences something like 12.8 years of subjective time between each pair of your keystrokes.

I originally conducted this thought experiment to try to illustrate the speed of computers in terms that could be understood intuitively. If you really wanted to be picky, however, you could reasonably argue that clock cycles aren’t really equivalent to heartbeats. So maybe as an alternative we should use instructions-per-second rather than cycles-per-second as our heartbeat metric. But even if we assume the processor averages a thousand cycles per instruction (which is probably an enormous overestimate, even for a modern short-cycle RISC CPU), the 2.8 GHz machine will still wait something like 336 subjective days between keystrokes, and that’s the lion’s share of a year. The old 6502, whose instructions were typically 5-6 clock cycles, would still have waited on the order of 7.3 subjective hours per character for your letter. It’s still a really long time, compared to our own human experience of the process.

1. Sources vary on this point; this range comes from Louise Simmers, Diversified Health Occupations, 2nd Ed., 1988, pg. 157. The 72bpm figure comes from Structure and Function of the Body, 9th Ed., 1992, pg. 231. Simmers gives a resting heart rate for infants in the 90-140 bpm range.

2. Source: National Vital Statistics Reports, Vol. 48, No. 18

3. The ratio is probably closer to 1.05:1 of males to females, at birth; but over age 65, the ratio drops to 0.72:1. So, let’s just call it 1:1 and be done with it.

4. Ignore, for the moment, the problem that a “real minute” does not really exist independent of your reference frame.

Filed by Michael at 13:46 under Miscellaneous, Philosophical
3 Comments

So, we recently learned that my advisor was granted tenure, and that is excellent news — both for Sean, and for those of us who are his students. University promotion is an example of the “up or out” theory of advancement, the kind that is apparently also used in the upper echelons of the military establishment: Either you do a good enough job to get promoted to a higher level, or you lose your job. You can’t stay put.

Needless to say, the process of acquiring tenure is among the most stressful, labor-intensive, and all-consuming experiences in a university faculty member’s life, so we thought it only fitting to help celebrate Sean’s clearing of this hurdle by playing some good-natured pranks on him. In the interest of full disclosure, I should say that “we” in this case means the graduate students in Sean’s lab, although I was only peripherally involved. The real perpetrators may be identified by the appropriately smug expressions on their scheming faces.

Suffice it to say that when Sean arrive to work on Monday morning, his office and the hall outside the laboratory had been turned into a kind of impromptu Viking village. His office table was removed, and replaced with a stone-lined firepit, over which was “roasting” a copy of the Orange Book and a couple of other choice publications on a wooden spit. I say “roasting” with quotation marks because, despite what you might think of us, we are not foolish enough to build a real fire in our advisor’s office. This one was constructed out of paving stones, tissue paper, logs, and Christmas lights. It was very realistic-looking, though. The verisimilitude of the Viking experience was supplemented by the presence of three hens, who spent the morning strutting around his office eating cracked corn and clucking a lot.*

There is, it turns out, a great deal of humour to the sound of a live cackling chicken in the halls of a computer science building.

Sean himself was crowned King of the Vikings with a horned helmet and a cloak, courtesy of a lot of last-minute papier mâché and clever sewing done by Scout. His wife and daughters turned up, appropriately attired and armed with a faux sword. A large collection of green and white balloons scattered through the hall added to the overall surreality of the scene. As Sean himself put it, once he got past being stunned, “This is definitely not the way Mondays usually go.” But, unfortunately, he did not take advantage of his Kingship to let us plunder the nearby laboratories for comfy furniture and free food.

Even so, a good time was had all around, and it was generally acknowledged that it was a morning well spent, despite the fact that none of us got any work done. Or perhaps because of that.

* No chickens were harmed, or even mildly traumatized, in the making of this particular prank.

Filed by Michael at 17:48 under Academic, Personal
6 Comments